Radioimmunotherapy with beta emitting radionuclides has demonstrated significant anti-cancer activity, but is limited by the long range, low potency and lack of specificity of the beta particles. As a consequence, except in radiosensitive lymphoid cancers, major responses can not be achieved easily without dose limiting myelosuppression or BMT rescue. As an alternative, targeted alpha particle therapy allows selective killing of single cells and small clusters of cells, but may not be effective in debulking large tumors. We have developed 2 markedly different forms of alpha therapy, which allows us to ask important questions about their use. Targeted Bi-213 is useful for targets within the vasculature. but is limited by its short 46 min halflife. Ac-225 atomic alpha generators, which yield a net of 4 alphas, have 10 day half-lives, allowing diffusion into bulkier tumors, but will likely be limited by toxicity from errant alpha emitting daughter products. Over the last 5 years, we have developed human therapeutic model systems for studying alpha radioimmunotherapy including novel in vitro chemistry and methods, animal models and human clinical studies. We hypothesize that by understanding the radiochemistry, cellular metabolism and catabolism and radiobiology of these radioconstructs in these systems, one can design clinical strategies to take full advantage of their unique and highly active features, while reducing their dose limiting characteristics. This will involve elucidating the role and the interrelationships of the following key parameters: Response rates and toxicity;tumor phenotype and genotype;target cell surface antigen density;radionuclide half-life and generator daughter cascades; single-cell, vasculature and tumor cell cluster geometry. These issues will be addressed in human clinical trials, in laboratory investigations associated with these trials, and preclinical model systems. First, we complete our validation of the concept that targeted alpha emitters can be safe and effective agents in acute leukemia (Aim 1). Next, we ask for the first time if targeted alpha-generators can be used in humans (Aim 2), and how can one control the possible errant daughter product toxicities (Aim 3). Finally, (Aim 4) we explore in a preclinical model whether alpha irradiation can be used to target the tumor neovasculature, using knowledge about the differences in targeted alpha emitters and alpha-generators, and propose optimal strategies for their use.